matrix-multiply commited on
Commit
bc093a5
·
verified ·
1 Parent(s): af63e26

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +67 -3
README.md CHANGED
@@ -1,3 +1,67 @@
1
- ---
2
- license: mit
3
- ---
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ language:
3
+ - en
4
+ - py
5
+ tags:
6
+ - code
7
+ - documentation
8
+ - python
9
+ - docstring
10
+ - dataset
11
+ license: mit
12
+ ---
13
+ # DocuMint Dataset
14
+
15
+ The DocuMint Dataset is a collection of 100,000 Python functions and their corresponding docstrings, extracted from popular open-source repositories in the Free and open-source software (FLOSS) ecosystem. This dataset was created to train the [DocuMint model](link-to-model-card), a fine-tuned variant of Google's CodeGemma-2B that generates high-quality docstrings for Python code functions.
16
+
17
+ ## Dataset Description
18
+
19
+ The dataset consists of JSON-formatted entries, each containing a Python function definition (as the `instruction`) and its associated docstring (as the `response`). The functions were sourced from well-established and actively maintained projects, filtered based on metrics such as the number of contributors (> 50), commits (> 5k), stars (> 35k), and forks (> 10k).
20
+
21
+ An abstract syntax tree (AST) based parser was used to extract the functions and docstrings. Challenges in the data sampling process included syntactic errors, multi-language repositories, computational expense, repository size discrepancies, and ensuring diversity while avoiding repetition.
22
+
23
+ ## Dataset Structure
24
+
25
+ Each entry in the dataset follows this structure:
26
+
27
+ ```json
28
+ {
29
+ "instruction": "def get_dataloaders(accelerator: Accelerator, batch_size: int = 16):\n \"\"\"\n Creates a set of `DataLoader`s for the `glue` dataset,\n using \"bert-base-cased\" as the tokenizer.\n\n Args:\n accelerator (`Accelerator`):\n An `Accelerator` object\n batch_size (`int`, *optional*):\n The batch size for the train and validation DataLoaders.\n \"\"\"\n tokenizer = AutoTokenizer.from_pretrained(\"bert-base-cased\")\n datasets = load_dataset(\"glue\", \"mrpc\")\n\n def tokenize_function(examples):\n # max_length=None => use the model max length (it's actually the default)\n outputs = tokenizer(examples[\"sentence1\"], examples[\"sentence2\"], truncation=True, max_length=None)\n return outputs\n\n # Apply the method we just defined to all the examples in all the splits of the dataset\n # starting with the main process first:\n with accelerator.main_process_first():\n tokenized_datasets = datasets.map(\n tokenize_function,\n batched=True,\n remove_columns=[\"idx\", \"sentence1\", \"sentence2\"],\n )\n\n # We also rename the 'label' column to 'labels' which is the expected name for labels by the models of the\n # transformers library\n tokenized_datasets = tokenized_datasets.rename_column(\"label\", \"labels\")\n\n def collate_fn(examples):\n # For Torchxla, it's best to pad everything to the same length or training will be very slow.\n max_length = 128 if accelerator.distributed_type == DistributedType.XLA else None\n # When using mixed precision we want round multiples of 8/16\n if accelerator.mixed_precision == \"fp8\":\n pad_to_multiple_of = 16\n\t elif accelerator.mixed_precision != \"no\":\n pad_to_multiple_of = 8\n\t\t else:\n pad_to_multiple_of = None\n\n return tokenizer.pad(\n examples,\n padding=\"longest\",\n max_length=max_length,\n pad_to_multiple_of=pad_to_multiple_of,\n return_tensors=\"pt\",\n )\n\n # Instantiate dataloaders.\n train_dataloader = DataLoader(\n tokenized_datasets[\"train\"], shuffle=True, collate_fn=collate_fn, batch_size=batch_size, drop_last=True\n )\n eval_dataloader = DataLoader(\n tokenized_datasets[\"validation\"],\n shuffle=False,\n collate_fn=collate_fn,\n batch_size=EVAL_BATCH_SIZE,\n drop_last=(accelerator.mixed_precision == \"fp8\"),\n )\n\n return train_dataloader, eval_dataloader",
30
+ "response": "Creates a set of `DataLoader`s for the `glue` dataset,\nusing \"bert-base-cased\" as the tokenizer.\n\nArgs:\n accelerator (`Accelerator`):\n An `Accelerator` object\n batch_size (`int`, *optional*):\n The batch size for the train and validation DataLoaders."
31
+ }
32
+ ```
33
+
34
+ ## Usage
35
+
36
+ This dataset was used to train the DocuMint model, a fine-tuned variant of Google's CodeGemma-2B that generates high-quality docstrings for Python code functions. For more information on the model and its training procedure, please refer to the model card.
37
+ Dataset Usecases
38
+
39
+ The DocuMint dataset can be used for various purposes related to code documentation and natural language processing tasks. Some potential usecases include:
40
+
41
+ Training and evaluating models for automatic docstring generation
42
+ Studying the characteristics and patterns of high-quality docstrings
43
+ Analyzing the relationship between code structure and its corresponding documentation
44
+ Developing tools for assisting developers in writing effective docstrings
45
+ Conducting research on the challenges and best practices in code documentation
46
+
47
+ Researchers, developers, and organizations interested in improving code documentation quality and automating the process of docstring generation can benefit from this dataset.
48
+
49
+ ## Citation [optional]
50
+
51
+ <!-- If there is a paper or blog post introducing the model, the APA and Bibtex information for that should go in this section. -->
52
+
53
+ **BibTeX:**
54
+
55
+ (TODO)
56
+ ```
57
+ @misc{poudel2024documint,
58
+ title={DocuMint: Docstring Generation for Python using Small Language Models},
59
+ author={Bibek Poudel* and Adam Cook* and Sekou Traore* and Shelah Ameli*},
60
+ year={2024},
61
+ }
62
+ ```
63
+
64
+ ## Model Card Contact
65
+
66
+ - For questions or more information, please contact:
67
+ {bpoudel3,acook46,staore1,oameli}@vols.utk.edu